Up to 584% faster EVM execution, with identical results.
NE3 is an execution engine for EVM chains: the same results, faster. Every figure on this page was measured on BSC mainnet, on real swaps and live quotes.
Every workload gets faster.
The gain grows with the share of execution spent running contract code rather than reading and writing state. The last three workloads are about the same size, sized at 37M to 40M gas, yet gain from 1.5x to 3.9x.
0 differences in 52,974 timed comparisons with the stock node.
Each of the 8,829 simulations was timed six times and checked field by field. Live swaps that revert by the time they run revert identically on both engines.
More work from every node.
A node running NE3 returns the same results as before, faster. Operators keep their chain and everything it computes.
Faster quotes
Users wait less for quotes. Builders and solvers try more candidates in the same time.
More simulations per server
Or the same load on fewer servers.
Opt-in, no hard fork
Each operator decides, node by node. The chain itself does not change.
Built for EVM chains
Today's figures are measured on BSC mainnet. NE3 is built for EVM chains in general.
Reproduce it against your own node.
We invite your engineers to check every figure. NE3 stays on our node, and your stock node is the baseline.
Run a stock node
A standard BSC node on a server like ours, under your control.
Run our measurement tool
It sends the same requests, on the same block, to your stock node and to our NE3 node.
Compare the results
Times come from each node's own log, cross-checked by the client. Results are compared field by field.
A few hours of runs, under a protocol agreed in writing beforehand.
Request the protocolRequest a benchmark.
Tell us which chain and which workloads matter to you. We reply with a protocol and a time slot.